In PtK1 cells anaphase onset is delayed until 23 +/- 1 min after the last monooriented chromosome acquires a bipolar attachment to the spindle--an event that may require r 3 hrs (Rieder et al., 1994. J. Cell Biol., 127:1301-1310). To determine the nature of this cell cycle checkpoint signal, and its site of production, we followed PtK1 cells by video LM after destroying specific chromosomal regions with the BMIRR's laser microsurgery system. The checkpoint was relieved, and cells entered anaphase, 17 +/- 1 min after both kinetochores and the centromere region on the last monooriented chromosome was destroyed. Thus, the checkpoint monitors an inhibitor of anaphase produced in the centromere region of monooriented chromosomes. Next, in the presence of one monooriented chromosome, we destroyed one kinetochore on a bioriented chromosome to create a second monooriented chromosome lacking an unattached kinetochore. Under this condition anaphase began in the presence of the experimentally-created monooriented chromosome 24 +/- 1.5 min after the non-irradiated monooriented chromosome bioriented. Thus the checkpoint signal is not generated by the attached kinetochore of a monooriented chromosome or throughout the centromere volume. Finally, we sought to destroy the invisible unattached kinetochore on the last monooriented chromosome. These cells either entered anaphase 20 +/- 2.5 min after the operation, without congressing the irradiated chromosome, or the cell delayed anaphase until the chromosome bioriented and moved to the metaphase plate. Correlative LM/3D EM of anaphase cells containing chromosomes that did not congress confirmed the absence of a kinetochore on the unattached chromatid. Together, our data reveal that molecules in or near the unattached kinetochore of a monooriented PtK1 chromosome produce an inhibitor of the metaphase-anaphase transition. This work was published in 1995 in the Journal of Cell Biology (see highlights)